Like the title says, I just bought a replica Apple 1 PCB and need a parts list. Ie: ICs, resistors, capacitors (with values if possible), etc.
My experience in this is admittedly limited. So having a full list would be more helpful than the schematics. I've repaired a number of vintage computers in the past. Mostly replacing chips and patching broken traces. Then moved on to piecing together what I like to call "frankenputers". Basically, I take various parts of dead machines. Put them together into some semblance of the machine I wanted. Then zap it full of electricity to bring my creation to life!
Now I'm in the process of building a ZX Spectrum. But this time, I bought a reproduction of the bare PCB and sourced all the parts to assemble it from "scratch". I'm almost done now, although I have no idea if it works yet. I won't find out until the RAM arrives in a few weeks. However, all the voltages are good on the board, so I'm hopeful.
Anyway, with this project almost complete, I figured I'd try something a little more challenging. Hence, the Apple 1. I've never used one before (never even seen one for that matter). But given the going price of a "real/original" computer, I figured that this is the only way I was ever going to get the chance.
To that end, I've been looking for a parts list. It was a serious help when building my Spectrum. I've tried Googling it, but no luck. I know that there are complete kits out there, but when building my Spectrum (or frankenputers for that matter), I've found that half the fun is finding all the little bits as I go along. Kinda like having a combination scavenger hunt and jigsaw puzzle. Anyway, if anyone has (or can point me to) one, I'd appreciate it.
Also, as a side question, and strickly a philosophical one at that. If Steve Wozniack were to assemble one of these boards, would that make it a real Apple 1? or would it still be just a replica? Just curious on that. Any thoughts?
Hi there,
Uncle Bernie, who is a very active member on this forum, sells his famous kit with burned-in ics on ebay: https://www.ebay.com/sch/uncle.bernie/m.html
His kits come with a link to a very helpful guide how to build your Apple-1 and what to watch out for. I can HIGHLY recommend his kits. I had almost no clue whatsoever about electronics when I bought a kit from him and still had been able to build my first Apple-1 replica.
Good luck!
Best
Armin
Everything you're looking for is described by Mike Willegal in his "Apple 1 Mimeo Computer Assembly
and Bring Up Guide". https://www.willegal.net/appleii/A1-assembly-v1.1.pdf
If I was you I'd just buy Uncle Bernie's parts kit and then consult Mike's document referenced above to fill in any missing pieces if there are any. Those guys have done a lot of the hard work for you.
In post #1, distantstar001 wrote:
"I've found that half the fun is finding all the little bits as I go along."
Uncle Bernie cautions:
Oh, seems we have found another daredevil maverick builder who loves to get defrauded by these Chinese counterfeiters and swinders, see here:
https://www.applefritter.com/content/warning-counterfeit-ds0025-china-oh-no-surprise
... so note that you have to dodge fake DS0025, fake 6502, fake 6520, fake 2504, fake or wrong 2513, fake 2519, fake dog food, fake everything, etc.
Of course, there are sources for these rare parts other than the crafty Chinese counterfeiters but the prices for the rare parts are staggering:
DS0025 - $40 (Rochester Electronics, IC Broker)
Signetics 2504V (date code 1977) = $17.99 (you need 7 --- google Unicorn Electronics Apple-1 parts)
Signetics 2519B (date code 1978) = $49.99 (some guys on Ebay want $65 ea plus shipping for 2519N)
Signetics CM2140 (date code 1978) = $49.99
4027 type DRAMs = $2.79 (you need 16)
LM323K = $71.17 (Digikey part# LM323KSTEELNS/NOPB-N)
Then you need to find all the TTLs (which are harder and harder to find) because LSTTL don't work in each socket of the Apple-1. And you need to buy the vintage carbon composition resistors in packs of 50 off Ebay. But you need only 1..12 per value.
Happy hunting !
- Uncle Bernie
Thanks! This is exactly what I was looking for!
As for the kits. I appreciate the advice (and I may eventually follow it). But for now, $300 is a bit much for me to be spending all at once. I realize that my current approach may cost me more in the long run, but then I am expecting this project to take a few months to complete.
The good news is that I do have a beat on at least one of the more obscure/expensive parts mentioned. Where I live, there's a local electronics store that carries the LM323K for about $13 (plus tax). I'm not sure if they've updated their prices since the '80s. I'll have to ask about the TTLs.
Question: I get that 74LS__s won't work, but what about 74ALS__ chips? From what I understand, they have the same speed and pinouts.
The 6502 is going to be an issue though. I have several other machines in need of them, including an unenhanced //e and a VIC-20. I've been using a 65c02 in the //e for the time being, but that won't work for the VIC. PolyPlay has them for about 11€ to 12€ (the later is a 6502B if that makes a difference). However, I understand that I could use an Motorola 6800 as well? But I haven't seen any examples of it online.
6502B would probably work in your VIC-20. There are adapters to use a WD65C02 in a VIC-20 also. Look up Byte Attic's "Pimp my 8 bit" Episode 2. 6800 in an Apple 1 is not really practical. There is a reason you don't see examples, there is no code for that. Plus 6800 chips are more rare and expensive than 6502s. NMOS 6502 aren't exactly rare. Millions and millions were made. They aren't really availabe NOS anymore but good working pulls from stuff being parted out can be had. The "right" ones for an Apple 1 may be harder to find. But if you're bodging things together piecemeal on a a budget you probably aren't in the market for the ceramic package early made ones that most people covet.
Funny thing is, a while back VIC-20s were the junk we'd buy off eBay to pull a 6502 out of to put in something else and then bin the rest of the VIC-20.
Actually looking at eBay VIC-20s are still available pretty cheap. If I was looking for a source for an NMOS 6502 I'd just get one of those or a 1541 (another place we'd pull 6502s from) off eBay and then toss the rest.
Thanks for this! I'll have to go over what I have, but knowing that some of my machines can handle a newer 6502 is a big help. That gives me 12 potential donors if I can't find an Apple-1 compatible.
Just out of curiosity, would a 6502B work in an Apple ][ Plus? I have 2 project boards that are more or less functional but don't have cases as of yet. I was going to mount one of them into a custom case with an external keyboard. Since it's never going to be original, I figure that it would be my best transplant candidate.
6502B is an NMOS 6502, it is just capable of 2MHz instead of only 1MHz like the original 6502. The Apple /// used a 6502B and ran at 2MHz for native /// app code, it slowed down to 1MHz for a lot of things including Apple ][/][+ emulation. It should work fine in a ][+, just obviously only at 1MHz there.
I can confirm that a genuine 6502B works just fine in the Apple-1 because this is what I provide with my famous IC kits.
They do have an advantage: the timing margins get wider and so the whole machine is more tolerant to errors in the 74123 timing.
But Woz' timing calculations were spot on: 480ns on the oneshot is the right compromise for the slower 6502 and MK4096 DRAMs of the time.
The only little problem is that you can't hit these 480ns unless you hand select the 74123 and the timing R's and C's.
I was able to make one of my Apple-1 builds run with a CMOS 6502 (W65C02S6TBG-14 from Western Design Center): all you need to do is to lift pin #1, #5, and #36 so they don't contact the IC socket and tie the lifted #36 to pin #8. You need added power supply bypass capacitors (a 100 nF on the solder side between pin #8 and pin #21), otherwise the fast switching 65C02 will cause trouble.
You can't use LSTTL, STTL or ALSTTL in all sockets of the Apple-1. I have populated one specimen with LSTTL but it does only work when the 2504 dynamic shift registers were MM1404 from National Semiconductors. All other known substitutes AND the original Signetics 2504 did not work in that LSTTL machine. High speed CMOS (74HC) is unfit to work in a TTL system and the 74HCT - which have the proper logic level thresholds - were designed for working in a TTL environment but again, in the Apple-1 they don't work (except for a few places).
The LM323K price problem came about because the last competitor (ST Microelectronics) stopped building them a few years ago, and only TI is left, so they can ask for moon prices. I was able to buy 100 of the "last call" LM323K made by STM so I'm set. People seem to be unable to understand that there are lots of drop-in substitutes like the UA323K made by Fairchild back in the day (before National Semiconductors swallowed Fairchild) and some types with extended temperature range like the LM123 in TO-3. You can still find plenty of those substitutes. Last time I looked, many moons ago, www.jameco.com had UA323K for about $10 but they are marked as "selling out", so once they are gone, they are gone. Or, as seen in post #6, you can often find those regulators in old electronic shops which existed since the 1970s. I was lucky to find dark brown 100nF bypass capacitors in one of these shops, which were close in looks to the BEL brand capacitors used in the Apple-1 1st production run sold to BYTE SHOP. Alas, these are gone now, I sold them out with my first kits and now regret I kept none for myself.
Still, despite of these occasional savings possible when finding rare parts in old electronics stores founded in the 1970s, I think it's impossible for individuals to put together the parts in my famous IC kits for less money than I ask for. There is no profit in it (by design, don't feed the parasites, "Who is John Galt ?") so you can't buy them cheaper elsewhere. And some of them are really heinously expensive since greedy IC brokers have figured out there is some small demand for the rare Signetics PMOS ICs used in the Apple-1. They sit on stockpiles from which many thousands of Apple-1 clones could be built, but their moon prices are so usurious that I predict this greed and usury will be a poison pill and it may put an end to Apple-1 cloning sooner than you think.
Uncle Bernie
Apple I clones with original 1970s chips are for collectors with deep pockets. This design is nearly 50 years old and several of the required components have not been made for 40 years. The layout is marginal and almost never worked at all, meaning that you cannot troubleshoot it without deep understanding of analog circuit principles. Collecting grab bags of chips that may or may not work at all, or may have the wrong analog characteristics, is a fraught endeavor.
74- series SSI chips with the same part number always have the same pinout, but that's the least of your problems. Each logic family (S, LS, ALS) has its own setup and hold timing requirements, slewing speed, drive strength, pin capacitance, and valid logic voltages (in other words, analog parameters). This can cause issues in the circuit because at that level all electronics is analog.
Woz designed the Apple I board to work with either the 6800 or 6502, which is historically interesting as it may be the only computer able to work with both. After designing the 6800 at Motorola in 1974, a large part of the engineering team left to form MOS Semi, which introduced the 6501 in 1975 as a pin-compatible (but not binary code compatible) alternative to the 6800. That prompted Motorola to file suit, and the settlement required MOS to change to a pinout that was not 6800 compatible, which was the 6502.
The features that enable the Apple I board to work with the 6800 include the clock driver area (unused with the 6502), the jumper matrix, some other jumper links, and the C1 socket for a 7404 inverter to generate the two-phase clock (which the 6502 does not need).
Nevertheless, the potential to use a 6800 was seemingly never tested or debugged since Woz did the first hand-wired prototype in 1975. The ROM monitor (Wozmon) and all the Apple I software is only compatible with the 6502. Finally in 2013, a Mimeo clone board was shown working with a 6800 and a ROM monitor translated to 6800 code:
https://www.youtube.com/embed/ag6pWUhps7U
It only worked by substituting some parts and adding pull-up resistors to certain processor pins, which were never documented, so the 6800 ability was apparently never tested by Apple at all. It's a fun thing to try, but none of the Apple I software (Basic, and the few dozen other programs) work with a 6800.
You wrote above that you can't use LSTTL in the Apple-1 replica. It's funny how sometimes people in completely different parts of the world get the same idea at the same time! About two or three weeks ago I decided to build an Apple II rev.0 replica, where almost all of the chips used were LSTTL. After searching in my den a bit I found some of them on the Apple II cards. And for some reason I decided to check which ones would work in the Apple-1 and which ones wouldn't. Here's a list of what I was able to successfully integrate: 3x74ls00, 74ls02, 74ls04, 74ls08, 2x74ls10, 74ls32, 74ls174, 74ls175 and 4x74ls257. Checked with Mike Willegal's 30-40 min test of each memory block and Integerm's Basic program download. No errors! I have AMI 1404 ARS shift registers from aliexpress on this board. As new LSTTLs become available for the Apple II rev.0 build I will continue this fun experiment.
In post #13, macintosh_nik wrote:
"You wrote above that you can't use LSTTL in the Apple-1 replica"
Uncle Bernie answers:
I wrote that you can't use LSTTL in ALL sockets of the Apple-1, meaning: don't stuff LSTTL in without a brain and expect it to work.
I did this to prevent frustration of builders who can't find TTL and use LSTTL instead.
The LSTTL was advertised to be a drop-in replacement for the TTL of the same type number, and if you look at the timing parameters it really seems they designed the LSTTL specifically to have the same timing as the TTL counterparts.
I also wrote here in Applefritter in many threads about my 100% LSTTL populated Apple-1 clone which works. But it needs National Semiconductors MM1404 to work.
I suspect it does not have to do with the timing, but with the Schottky clamping diodes on the inputs which do not allow the signal to swing much below GND (0V). And it seems that some of these PMOS shift registers need that to work. This may be a margin effect, too. Some process corners may work, others, not so.
So in the end, even if you have other PMOS shift registers than these elusive MM1404 (very hard to find !) you still can use LSTTL at many places in the Apple-1 but you can't do this with a broad brush, putting them in everywhere.
If I had more time to investigate this then a list could be made at which places LSTTL can be used without problems.
But this investigation should be done by volunteers. I just don't have the time to spend hours on this topic of LSTTL. I also have plenty of all the TTL types needed in the Apple-1, much more than the rarer Signetics PMOS ICs of the 2504, 2513, 2519 type (or their substitutes).
My main motivation to experiment with the LSTTL Apple-1 was to get current consumption down. With 100% LSTTL it is low enough to use original National Semiconductors LM309K regulators from which I have a few with the super rare "golden base" TO-3 package. The industry quickly abandoned them for the "Steel" base versions, so "golden base" versions are exceedingly rare. But the original Apple-1 had "golden base" regulators. Another tell that it's a clone. It is impossible to fake the "golden base" / put gold paint on a "Steel" base, it always looks wrong. But it is possible to put a "LM323K" with the correct logo or date code on a real "golden base" LM309K. Of course, all the 74LS would also need re-stamping to plain vanilla TTL, but this would be necessary anyways, for the "right" manufacturer logos and date codes. Such a chip set could be used to restore originals and command a high price. But in a clone it's useless, the PCB is the "tell" and no amount of money or effort can reproduce these original PCBs.
These were my motivations back in the day why I looked into the LSTTL substitutes. But now, years later, with a TTL shortage being noticable, it might be worthwhile to work out a plan for LSTTL substitutions. I always thought that we would run out of 2504/2513/2519 before running out of TTL and then the huuuge stocks of these 2504/2513/2519 were discovered (although at usurious prices). But TTL is getting scarce. Not all of them, of course. Still, we should have a Plan B. LSTTL could be part of that plan.
- Uncle Bernie
So I've begun my parts search. The good news is that the electronics store I mentioned did have all but a handful of the TTL chips I need, almost all with date codes ranging from 73 to 77 (not that important, but a nice surprise). Unfortunately, that was about all they had. The shift registers and DRAM are going to be hard. They may have them, along with the character generator, but they're going to have to dig for them.
I also managed to find a second parts list that was easily copied into a Numbers spreadsheet so that I can check off parts as I find them. However, there is a complication. For the most part, the two lists match, except for one thing. The new list states that a 7451 was used where the board sights a 7450.
I've tried looking at reference pictures to resolve the discrepancy. However, there are three issues with that approach. First, I can only find three images with a high enough resolution to read the chip numbers on the boards. Two, none of these boards are being shown in operation, thus no way to know if they work or were just for show. And three, the pictures don't agree. Two had 7451s while the third had a 745o. So which is it? 7450 or 7451?
Original boards usually sport a 7451 in that location, but either part number will work.
Mike Willegal
In post #16, Mike Willegal wrote:
"Original boards usually sport a 7451 in that location, but either part number will work."
Uncle Bernie adds:
even 74LS51 will work ;-)
Some people think it's the most difficult to find TTL in the Apple-1. Sigh !
Rest assured, I realize that the TTL is the low-hanging fruit. The ROMS, the RAM, and the Shift registers are going to be the really hard parts! Although, not necessarily in that order.
In post #18, DistantStar001 wrote:
"I realize that the TTL is the low-hanging fruit. "
Uncle Bernie says:
Most TTL are easy to find, especially if you have a "vintage" electronics shop nearby which was founded in 1970s or early 1980s.
Some TTLs are hard to find in larger quantities. The 74161 is the worst. Those who still have them want moon prices.
"The ROMS, the RAM, and the Shift registers are going to be the really hard parts ! "
Uncle Bernie says:
None of these are hard to find. I gave you some sources in prior posts. The problem is the moon prices being asked for these parts nowadays. It's not the shops like Unicorn, mind you, it's the greedy and usurious IC brokers who are the culprit. These long obsolete PROMs, PMOS shift registers, and early 4k x 1 DRAMs were sitting in warehouses for 40+ years and did not move. And then, suddenly, there is some demand thanks to the Apple-1 cloning crowd (a small crowd worldwide, and the only "market" for these ICs). And so these greedy bastards jack up the prices every time somebody asks for these Apple-1 specific parts. During my quest for these parts I have discovered huuuuge stocks of these rare ICs from which we could build thousands of Apple-1 clones. But the IC brokers sitting on them got so greedy that they have poisoned the market. And they pissed me, the only quantity buyer in the world for these parts at that time, off in a big way. When the same usurious guy who sold me the 2919N for less than $17 three years ago suddenly wants $65 each, he can keep them, and wait for 200 years he will need to sell them piecemeal on Ebay (he sits on 5000 of them), and I keep track of the number he sells. It's pathetic ! He sells maybe a dozen per year. I bought more than 100 from him, before he got too greedy. Now I would not buy any from him anymore even if he would sell them to me for the previous price. It's over for him.
So, after my current stock of ICs runs out, no more of my famous Apple-1 IC kits anymore. Because I don't support usury. I have no issues when small businesses add some reasonable profit margins to keep their doors open, but there is a thin red line these usurers have crossed. By far. I'd recommend anyone building Apple-1 to boycott these usurers and tell them they you don't accept their usurious prices. ALL of the Signetics PMOS ICs in the Apple-1 can be replaced with far cheaper substitutes. So those guys who sit on more than 6000 of the Signetics 2504 can keep them. Other than then Apple-1 cloner scene there are no takers.
I had the same situation with 256x4 PROM blanks I found in Sweden, more than 1000 pcs there. Nobody other than the Apple-1 crowd would want them in quantity (even in the arcade machine scene these are rarely needed). These greedy bastards were no willing to come down with their moon price even to the level of 512x4 PROMs. Their demands were ridicolous: twice the money for half the memory, per chip. It should have been half the price of the 512x4 ones.
So I bought 512x4 PROMs from another source and added the extra bonus pages: the diagnostics page in the A1, A2 PROMs and the "extended format" page in the ACI PROMs. The first page in my 512x4 PROMs of course is the unmodified Woz code (used with permission from Woz himself !).
Other people have developed cheap substitutes for the 2519 based on six CMOS shift registers and we have explored drop-in for the 2513 character generator and for the 2504. We don't need to yield to the extortion from these usurers. Shall they go to hell !
Comments invited !